Tim Prince wrote: > bootstrap failures are due to a broken #ifdef specific to cygwin in the > headers provided with cygwin,
If you mean the strsignal change in libiberty, that's been fixed in CVS for a long time. > the requirement for a specific version of > autoconf (not available in setup), You are never supposed to need any version of autoconf to simply build gcc, only if you need to rebuild generated files. If the timestamps are correct (either by using a tarball or contrib/update_gcc to update from SVN) then autoconf should never be invoked, otherwise something is wrong with your setup. > and the need to remove the -werror in > libstdc++ build (because of minor discrepancies in cygwin headers). All That is what --disable-werror is for. > be accepted. Support for 64-bit native looks like it will be limited to > mingw, so I no longer see a future for gcc on cygwin. I don't know where you keep getting this conjecture from. Porting the Cygwin runtime library itself to x64 is a lot more work than getting a 64 bit MinGW environment working because Microsoft has already done the work in the case of a 64 bit MSVCRT, and because there is a lot more going on in the Cygwin runtime than in the Microsoft. But just because that porting work hasn't happened yet doesn't mean it won't eventually. Moreover, from the standpoint of gcc the two targets share virtually all the same code and there is very little substantial difference between them aside from miscellaneous specs file differences and driver bits, so once that porting of the Cygwin runtime happens there should be minimal work left on the gcc side. Brian
